Everyone knows this is a trap, and in trying not to fall in, most people fall in a different way. They pick a mistake so small and so safely resolved that it is obviously chosen to be harmless. "I once sent an email to the wrong person, but I caught it and recalled it." The interviewer hears the calculation behind it and mentally marks you down, not for the mistake, but for the dodge.
The real risk of this question is not admitting a fault. It is looking evasive. A candidate who can describe a genuine mistake plainly, without spiralling into self-flagellation or minimising it into nothing, signals something valuable: they can be trusted to tell you when something has gone wrong, which is exactly what you want from someone you cannot watch every minute.
What the interviewer is really testing
Not "do you make mistakes". Everyone does, and they know it. They are checking three things:
- Can you own it cleanly? Without excuses, without blaming the situation, and without the theatrical over-apology that is really just a bid for reassurance.
- Did you handle the fallout well? The mistake matters less than what you did once you realised. Did you flag it, fix it, contain it, or hide it and hope.
- Did you actually change something? A mistake you learned nothing concrete from is just a mistake. A mistake that led to a specific change in how you work is evidence of judgement.
A worked example
"In my first proper job, in a small finance team, I was reconciling supplier payments and I misread a figure, approved a payment for the wrong amount, about four hundred pounds too much to a supplier. I did not spot it at the time. I found it two days later when I was checking the week's runs.
My first instinct, honestly, was dread, and the tempting thing was to quietly try to claw it back and say nothing. But I went straight to my manager and told her before she found it herself, because a small error I had flagged was a very different thing from one she discovered.
We contacted the supplier, who were fine about it and returned the overpayment, and it was sorted within the week. The bigger thing was what I changed afterwards. I had been checking figures once, quickly. I started doing a second pass on anything over a threshold, on a different screen so I was actually re-reading rather than confirming what I expected to see. I never made that class of error again.
What it taught me was that the mistake itself was recoverable in about ten minutes of admitting it, and the only way to make it a real problem would have been to hide it."
Why it works
The mistake is real and has a number attached to it, four hundred pounds, which stops it feeling chosen for safety. But the answer is not about the mistake. It is about the fork in the road: the moment the candidate could have hidden it and instead told their manager first. That decision is the whole point, and they name the reasoning ("a small error I had flagged was a very different thing from one she discovered"), which shows they understand how trust works.
Then there is a specific, mechanical change, a second pass on a different screen, not a vague "I learned to be more careful". Anyone can promise to be more careful. Describing the exact new habit proves the lesson took. And the closing line reframes the entire event as a lesson about honesty rather than about accuracy, which is the more impressive read.
The ways this answer fails
The fake mistake. A mistake that was instantly caught, harmed nothing, and reflects no real fault is transparently defensive. Interviewers have heard "I'm a perfectionist" a thousand times. Pick something that actually cost something.
The excuse-laden version. "I made a mistake, but really it was because I had not been trained properly and the system was confusing and my manager had not..." The moment you start distributing blame, the answer collapses. Own your part fully, even if others contributed.
The over-apology. Do not perform how terrible you felt for three sentences. It reads as fishing for reassurance and it wastes the time you should be spending on what you did. A clean "I got it wrong" is stronger than a wounded one.
No change at the end. If the story stops at "and we fixed it", the most important part is missing. The specific thing you now do differently is what turns a confession into evidence.
A catastrophic, disqualifying mistake. Use judgement about scale. A mistake that suggests a fundamental lack of care or a serious ethical lapse is not the one to choose. Aim for a real error with a recoverable outcome.
Choosing yours
The best mistakes to tell are ones where you caught or owned the fallout well, and changed something concrete after. Look for a moment you genuinely got something wrong and then handled it like an adult. Jobs, projects, and volunteering all supply these. The failure question is a close cousin and the same principles apply, though "mistake" usually wants something smaller and more specific than "failure".
The hardest thing to control here is tone: too breezy and you seem to not take it seriously, too heavy and you seem fragile. Level and matter-of-fact is the target. Say it aloud and listen to where you land. You can practise it and hear how it comes across, which is far better than discovering your tone is off halfway through the real thing.